Claude Code Mods: My Top 3 and the Full List of 6,000+

Mods need Claude Code 2.1.287 or later, and they're on by default from there. Check your version with the first command below. If it's older, the second one updates it.

Mods work in the terminal, in the Code tab of the Claude desktop app and in the VS Code extension. After you install a mod, run /reload-plugins or start a new session so it loads.

A mod is a plugin that runs inside Claude Code itself. That's the big difference from a skill or an MCP, which only give Claude instructions or tools.

A mod can draw a pane next to your chat or a little stage above the prompt. It can add new slash commands, and it can step in right before Claude runs a tool. You install one the same way as any plugin, with /plugin. If you want the deep dive, Anthropic's mods docs cover all of it.

If you want skills too, these are the 5 free ones I use for front end work.

hcom connects your coding agents so they can message each other. Say you have two Claude Code sessions working on the same project. Now one can drop a message straight into the other's chat, and if that session is sitting idle, the message wakes it up and it gets to work.

It doesn't stop at two. An agent can spin up a whole team of agents, split the work and collect the results. It also works with Codex, Gemini CLI, Cursor CLI and a lot more, so all of them can talk to each other.

hcom installs as a small command line tool. The first command below uses uv. On a Mac you can run brew install aannoo/hcom/hcom instead, and the hcom page has one-line installers for Windows and Linux. Then start Claude Code through hcom with the second command. Running hcom on its own opens a dashboard where you can watch every message.

Open two terminals in the same project and run hcom claude in both. Then paste this into the first session. It's the moment from the video, one session handing work to the other.

Clawd Actor puts a little stage above your prompt. While Claude works, Clawd walks on and acts out whatever Claude is doing right now. There are 32 scenes, and they change as Claude moves from one job to the next.

When Claude reads through your files, Clawd turns into a detective with a fedora and a magnifying glass. When Claude spins up sub-agents, Clawd becomes a cowboy herding sheep. Editing code makes it an artist, and MCP tools turn it into a wizard.

You can play with it too. /clawd pet, /clawd feed and /clawd dance all work even while Claude is busy. Type /clawd with anything else and it answers in a speech bubble, which uses a tiny bit of your usage.

Run the three commands below inside Claude Code, send a prompt, and boom, Clawd is on stage. It's a fan-made mod, not an official Anthropic one. It shows in the terminal and the desktop app's Code tab, but not in the VS Code chat panel.

Flightdeck puts a live dashboard right inside your terminal. You see every agent Claude has running and what each one is working on. Each agent gets its own card with its context size and a running clock.

It also shows your context meter and your session cost. Every permission check shows up as a colored tile, so you can see exactly what got allowed and what got denied. Every number comes from your own session and nothing leaves your machine. It only watches, so it never slows Claude down.

Type /flightdeck to open it and /flightdeck close to hide it. On a wide terminal it can open by itself when a session starts.

📚 Open the full list of 6,000+ mods

This is the list from the video. awesome-claude-code-mods scans GitHub for every Claude Code mod and lists over 6,000 of them. A new mod usually shows up within a few hours.

The best part is it shows what each mod can access before you install it, so you know what it can read, write, run or send over the network. If you'd rather search than scroll, the same list lives at mods.aidojo.si. You can search by name or by the job you want done, and filter by access.

It's an independent scan, not an official Anthropic directory, so still read a mod's page before you install it.

Mods run with your permissions. A mod can read and write your files, run programs, make network requests and read your API keys. Mods aren't sandboxed, so only install them from people you trust.

Check what it does first. Clone the mod and run the command below on its folder. It lists every event the mod listens to and everything it asks Claude Code to do, without running any of it.

Turn one off anytime. Disable a single mod in the Installed tab of /plugin. To run one session with no mods at all, start Claude Code with --safe-mode.

Anthropic shares sample mods of its own in the claude-code-playground repo.

token-weather draws a forecast of your context window above the prompt.

blast-radius holds a risky command like rm -rf or a force push and shows you what it would change, with buttons to go ahead or cancel.

replay-theater adds a /replay command that steps through every edit Claude made in the last turn.

To try one, clone the repo and start Claude Code with the mod's folder, like claude --plugin-dir ./claude-code-playground/claude-code/mods/blast-radius. That loads it for one session without installing anything.

You don't need to write any code to get a mod. Claude Code has a built-in skill for building them, so all you need to do is describe what you want. Paste this, or swap in your own idea.

Claude asks you to approve each file it writes. When it asks about hot reloading, pick Enable for this session, and the mod loads the moment Claude finishes. If you want it different, just tell Claude what to change.